IBM Communications Server for AIX, V6 New Features and Implementation Scenarios

118 IBM Communications Server for AIX, V6: New Features and Implementation Scenarios

Technically, a Branch Extender poses as a network node with DLUR support

to computers or 3270 terminals in its local branch. At the same time it poses

as an end node to computers in the WAN. By impersonating an end node, a

Branch Extender gains the following advantages:

• It does not receive topology data or directory broadcasts from the wide

area network. Instead, it lets a network node in the WAN choose routes for

computers in the branch. A Branch Extender relays data between the

branch and the WAN and enables HPR traffic to flow to computers in the

branch.

• It can register all the branch’s resources to a directory server in the APPN

backbone network so that no manual definitions are needed there.

Network efficiency is improved by notifying the central directory server of

each resource’s location, to greatly reduce or even eliminate broadcast

searches in the WAN.

Branch Extender also allows direct, peer-to-peer communication between

branches connected to the same APPN connection network (CN). Though a

connection network configuration is not required, Branch Extender provides

substantial benefit when the uplink and downlink networks are CNs, for

example a branch CN such as token-ring, Ethernet or FDDI connected via a

BrNN to a WAN CN such as ATM.

7.2 Topology databases

APPN networks consist of a backbone structure of network nodes

interconnected by TGs, known as intermediate-routing TGs, and TGs

connecting end nodes to adjacent network nodes, virtual routing nodes, or

other end nodes, known as endpoint TGs.

Information about the backbone structure of the APPN network is kept within

the network topology database, which resides on every APPN network node.

Information about endpoint TGs is contained within local topology databases,

which reside on every APPN node or LEN end node.

The primary use of local and network topology databases is to enable route

calculation when an LU residing in one APPN node wishes to establish a

session with an LU residing in another APPN node. The topology databases

enable topology and routing services in the network node (NN) server to

determine all possible routes between the nodes. The local topology

database contributes the end node’s TG, while the network topology

database supplies the information on network nodes and the TGs between

them.

Chapter 7. Branch Extender 119

7.2.1 Local topology database

The local topology database contains information on all of the transmission

groups (TGs) attached to the node. Every APPN node maintains a local

topology database. An APPN end node uses the information in its local

topology database to send local TG information to its network node server in

locate requests and replies. In a network node, the local topology database

includes information about the attached end nodes. In HPR networks, the

APPN end node’s TG information (carried in TG vectors) indicates whether

the nodes are HPR-capable.

The local topology database is created and maintained by the end node

Topology Database Manager (TDM). It is not saved across reboots and is

rebuilt when the node initializes. Entries in the local topology database are

created automatically when configuration services informs TDM about newly

activated or changed TGs. The operator updates the local topology database

through configuration services. The local topology database is searched by

TDM when it receives a query from route selection services or from session

services.

7.2.2Network topology database

Network nodes in an APPN network need to maintain topology information

about the location of other nodes in the network and the communications

links between them, and to forward this information around the network when

the topology changes. The information is maintained in a database called the

network topology database. As the network grows in size, the amount of

stored information and topology-related network traffic can become large and

difficult to manage. Topology database size is a function of the number of

NNs and links, as well as the frequency of link state changes.

Every APPN network node maintains a network topology database in addition

to its local topology database. The network topology database does not

include information on APPN end nodes, LEN end nodes, or the transmission

groups attached to them. It includes information only on network nodes in the

APPN network and the transmission groups interconnecting them.

Network nodes in an APPN network send one another topology database

updates (TDUs) over CP-CP sessions whenever a resource (network node or

a TG between network nodes) is activated or deactivated, or its

characteristics change. Only the current changes are included in the TDU.

Every network node receives the TDU containing the current change, so each

has the same view of the network. The network topology database is used by

the network node to select routes for sessions that originate at the LUs in it

and at the end nodes that it serves. Unlike the directory database, which is